Republic of Tunisia ‫الجمهور ية التونسية‬

Ministry of Higher ‫وزارة التعليم العالي‬

Education and Scientific ‫والبحث العلمي‬
Research

Private Higher School of ‫المدرسة العليا الخاصة‬

Aeronautics and ‫للطيران و التكنولوجيا‬
Technologies

Final Year Project

Second Year, Engineering Cycle in Aeronautical

Titre du projet:

STUDY AND IMPLEMENTATION OF A CHAOTIC CIRCUIT

Author:
Amin BELLAAJ

Supervisor:
Dr. Abir LASSOUED

Academic Year 2024/2025

ABSTRACT

This report presents a study and an implementation of a chaotic circuit (Chua’s circuit) applied to secure data
transmission, particularly in military systems such as aircraft. The project explores the theoretical
foundations of chaos, emphasizing the properties of the Chua system, including sensitivity to initial
conditions, strange attractors, and positive Lyapunov exponents. A detailed analysis of the circuit’s dynamic
behavior was conducted using simulations in MATLAB-Matcont and Multisim, validating its ability to
generate stable chaotic signals (double-scroll attractor). The practical design of the circuit, based on
conventional electronic components (gyrator, Kennedy diode), was realized, followed by master-slave
synchronization tests to enable signal recovery by an authorized receiver. This synchronization, achieved
through unidirectional coupling, demonstrates the effectiveness of chaotic systems in masking and
transmitting data securely, opening up promising prospects in cryptography.

Keywords: Chaotic systems, Chua’s circuit, Strange attractors, Synchronization.

Table of Contents
List of Figures...........................................;;........................................................................................... 5

General Introduction............................................................................................................................. 7

Chapter 1: State of the Art ................................................................................................................... 8

1.Introduction .......................................................................................................….......................... 9

2.Secure Communication for Military Aircraft Data Transmission......................................... 9

3.Secure Communication Based on Chaotic Systems................................................................ 16

4.Signature of Chaotic Systems .................................................................................................... 19

5.Problem Statement........................................................................................................................ 21

6.Conclusion...................................................................................................................................... 22

Chapitre 2: Dynamic Behavior Analysis of the Chua System ......................................... 23

1.Introduction ................................................................................................................................... 24

2.Modeling and Stability Study of the Chua System................................................................... 24

3.Study of Equilibrium Points and Their Stability....................................................................... 27

4.The Chua Strange Attractor.......................................................................................................... 32

5.Richness of Dynamic Behavior Beyond Chaos....................................................................... 38

6.Conclusion...................................................................................................................................... 40

Chapitre 3:Implementation of the Chua Circuit and Synchronization................................. 41

1.Introduction ...........................................................................................................;....................... 42

2. Multisim......................................................................................................................................... 42

3.Design of the Chua Circuit.......................................................................................................... 42

4.Description of the Different Components of the Circuit .................................................….... 44

3
5.global Chua Circuit....................................................................................................................... 48

6.Observing Simulation Results in Multisim................................................................................ 49

7.Experimental Validation of the Chua Circuit............................................................................. 52

8.Synchronization............................................................................................................................. 55

9. Implementation of Coupling via Voltage Follower on the Chua Circuit............................. 57

10.Conclusion.................................................................................................................................... 60

General Conclusion............................................................................................................................... 61

Bibliography ...................................................................................................................................... 62

Appendix................................................................................................................................................. 63

4
List of Figures
Figure 1. Data transmission diagram………….............................................…..…...……..........10

Figure 2. Analog and digital signal transmission ……….............................................................10

Figure 3. Sampling and ADC (Analog-to-Digital Converter) representation……...…................11

Figure 4. Communication system modeling…….......................................………….....…........12
Figure 5. Parallel transmission diagram……....................….........................……….....…..…....13
Figure 6. Serial transmission diagram…..............….........................……….....….......................13
Figure 7. Transmission mode diagram….......................….........................…..............................14
Figure.8. Asynchronous Transmission of N Bits..........................................................................14

Figure 9. Synchronous transmission of N bits….....................................….........................…....14

Figure 10. Secure transmission diagram using two Colpitts oscillators...................................…17
Figure 11. Chua’s circuit……………........................….........................……….......…...............19
Figure 12. Asymmetric attractors and spiral attractors….........................................................…20

Figure 13. Chua’s circuit 2………........................….........................……….....…......................24

Figure 14. Graphical representation of a function h(x)........................……….....…....................25
Figure 15. Double-scroll attractor in phase space x-y-z.................…….….....…........................33
Figure 16. Double-scroll attractor x-y.....................….........................……….......…..................33
Figure 17. Double-scroll attractor x-z....................….........................……….......…...................33
Figure 18. Double-scroll attractor z-x.....................….........................……….......…..................34
Figure 19. Sensitivity to initial conditions for variable x (blue curve corresponds to IC1 and red
curve to IC2)........................................................….........................……….....…........................36
Figure 20. Sensitivity to initial conditions for variable y (blue curve corresponds to IC1 and
green curve to IC2)........................…............................……….....…...........................................36
Figure 21. Sensitivity to initial conditions for variable x (blue curve corresponds to IC1 and red
curve to IC2).......................…............................……….....…......................................................36
Figure 22. Stable dynamic regimes........................…............................…….....…......................39
Figure 23. 3D limit cycles........................…............................……….....…................................40
Figure 24. 2D periodic oscillation (x,y)..................…............................………..........................40
Figure 25. Simplified Chua’s circuit............…..........................……….....…..............................42
Figure 26. Assembly of a negative resistance........................……….....…..................................45

Figure 27. Characteristic i=f(U) of a negative resistance....….....................................................45

Figure 28. Kennedy diode........................…...........................................………...........…...........46

5
Figure 29. Characteristic i=f(U) of a Kennedy diode......….…….....….......................................46
Figure 30. Antoniou gyrator implementation to simulate an ideal inductance.............................48
Figure 31. Assembly of Chua’s circuit on Multisim...............................………..........…............49

Figure 32. Limit cycle (periodic regime) for R=200............................……….....…...................50

Figure 33. Rössler attractor (a) for R=190 and (b) for R=191..............……….....…...................50

Figure 34. Double-scroll attractor (a) for R=189 and (b) for R=170...……….....…....................50

Figure 35. Double-scroll attractor (a) for R=175 and (b) for R=155....……….....…...................51

Figure 36. (a) Main (b) Delayed for R=175……........................................….....….....................51

Figure 37. Hysteresis (a) for R=175 and linear line (b) for R=155.........….....…........................51
Figure 38. Assembly of Chua’s circuit..........….......................................………..........…...........52

Figure 39. (a) and (b) Limit cycle (periodic regime)................................……….........…...........53

Figure 40. Rössler attractor..................................……….....…...........……….....…....................53

Figure 41. (a) and (b) Main double-scroll attractor....……….....................…..............................53

Figure 42. (a), (b), (c), and (d) Double-scroll attractor….....…...........……….....…....................54

Figure 43. (a) Hysteresis and (b) Linear line.......……….....…...........………........….................54

Figure 44. Unidirectional synchronization..........……….....…...........……….....….....................55

Figure 45. Bidirectional synchronization.........……….....…...........……….....…........................56

Figure 46. Synchronization via follower circuit for Chua’s circuit.....…….....…........................57

Figure 47. Follower coupling of Chua’s circuit…….......…...........……….....….........................58

Figure 48. Synchronization circuit for Chua......…….....…...........……….....…..........................58

Figure.49. (a) Curve for the first circuit �1 = � � and (b) �2 = � � for the second circuit......59

Figure.50. (a) and (b) Double-scroll attractor of the first and second circuits.....….....................59

Figure.51.Superposition curve of two signals (Synchronization).....…........................................59

Figure.52. (a) Case of no synchronization and (b) case of synchronization.................................60

6
General Introduction

In a world increasingly interconnected, modern defense systems, such as military aircraft,

exchange critical data such as tactical coordinates, mission orders, or radar data, requiring
absolute protection against any interception or alteration. However , classical encryption
methods show their limitations in the face of the evolution of sophisticated cyberattacks and the
growing power of quantum computers, unable to guarantee long-term security. It is precisely
here that chaotic systems, known for their unpredictability and dynamic complexity,
emerge as an innovative solution for secure information transmission. Their ability to
generate pseudo-random signals, combined with unique mathematical properties such as
sensitivity to initial conditions and strange attractors, offers a promising alternative to hide
sensitive data and make it virtually undecipherable without a synchronized receiver.

Among these systems, Chua's circuit, introduced in 1983 by Leon Chua, It is based on three main
axes: the in-depth study of the circuit's dynamic characteristics, including its chaotic behavior
and unstable equilibrium points, the practical realization of the latter using standard electronic
components (resistors, capacitors, operational amplifiers) to validate its operation, and finally the
testing of a master-slave synchronization system allowing a receiver to recover the original
message by aligning its own chaotic circuit with that of the transmitter.

This report is structured into three chapters. Chapter 1 presents an overview of chaotic systems
and their application to secure data transmission, highlighting the challenges of communication
in military environments. Chapter 2 analyzes in detail the dynamic behavior of Chua's circuit,
including its mathematical modeling, the study of its equilibrium points, the characterization of
its strange attractor (double-scroll), and the demonstration of its chaotic properties via
simulations using MATLAB-Matcont. Chapter 3 describes the practical implementation of
Chua's circuit using standard electronic components (gyrator, Kennedy diode), simulation results
using Multisim, and the implementation of a master-slave synchronization system to validate its
ability to transmit data securely.

7
Chapter 1: State of the Art

8
Chapter 1: State of the Art

1. Introduction
Chaos, a fascinating and complex phenomenon, manifests itself in a wide variety of nonlinear
dynamical systems. It represents a boundary between order and disorder, where seemingly
random behaviors emerge from deterministic laws. This chapter explores the theoretical
foundations of chaos, focusing on key concepts such as dissipativity, strange attractors, and
sensitivity to initial conditions. These properties open up practical perspectives in fields such as
secure communication.

2. Secure Communication for Military Aircraft Data Transmission

In military environments, secure communication is crucial for protecting sensitive information.

Military aircraft exchange critical data such as mission orders, target coordinates, and radar
information, requiring robust solutions to prevent interceptions and cyberattacks.

The main challenges in this type of communication are:

 Confidentiality: Preventing adversaries from intercepting messages.

 Authenticity : Ensuring that received data comes from a trusted source.
 Availability : Guaranteeing continuous communication even in the presence of
interference .

2.1. Transmission des données

2.1.1. Définition
The transfer of digital information represents the movement of data from a source to a recipient,
using a physical medium called a communication vector. This medium can take various forms
such as metal cables, glass fibers, or wireless waves. Figure 1 shows a simplified diagram of the
data transmission process in a digital communication system.

This technical process enables the routing of encoded information from its origin point to its
final destination. It can concern both digital signals and analog signals.

9
Chapter 1: State of the Art

Figure.1. Data transmission diagram

2.1.2. Operating Process

2.1.2.1. Data Encoding

Original information, whether in the form of text, audio, or video, is converted into signals
suitable for transmission. This is typically done by transforming the data into binary code (0s and
1s) or other formats understandable by communication protocols. Digitization aims to convert a
continuous analog signal into a discrete signal composed of a series of defined and limited values.
This process allows the representation of the initially continuous information as exploitable
digital data in binary form. Figure 2 illustrates the process of converting an analog signal into a
digital signal.

Figure.2. Analog and digital signal transmission

The transformation of an analog signal into a digital signal occurs in two main stages: sampling
and quantization (or analog-to-digital conversion). During sampling, it is crucial to select an
adequate number of samples. This number must be sufficiently high to faithfully reproduce the
original signal and capture all its essential variations. However, care must also be taken to avoid
an excessive number of samples, which would make the file overly large and complex to process,
potentially generating unnecessary storage and computational overhead. Thus, the goal is to
strike an optimal balance between representation accuracy and practical efficiency.

10
Chapter 1: State of the Art

Figure.3. Sampling and ADC (Analog-to-Digital Converter) representation

2.1.2.2. Signal Transmission

The encoded data is sent via a transmission medium, which can be a wired connection (such as
optical fiber or coaxial cable) or a wireless connection (such as radio waves).

2.1.2.3. Propagation through Networks

The data travels through a network infrastructure composed of routers and switches, which direct
the data toward its destination. Devices may be used to amplify or reprocess the signal to ensure
clarity over long distances.

2.1.2.4. Reception

At the destination, the receiving device decodes the signal to convert it back to its original format.

2.2. Modeling a Communication System

2.2.1. Definition

 Transmitter Equipment This is a device or component that sends information through a

communication channel. The latter adapts the data into signals suited for transmission,
whether expressed in continuous or discrete form.

 Transmission Medium This medium carries the physical signal from one point to
another. It can take various forms, such as specialized cables (coaxial or twisted pair),
wireless transmissions (radio waves or infrared), or optical fibers.

 Receiving Device This equipment captures the signals from the transmitter and interprets
them to recover the original data. To ensure accurate data recovery, it must be
coordinated with the transmitter. Figure 4 represents a simplified model of the
communication process.

11
Chapter 1: State of the Art

Figure.4. Communication system modeling

2.2.2.2. Transmission Method

The data transfer method defines the volume of basic information units (bits) that can be
transmitted simultaneously through a communication medium.

In an interaction between two devices over a communication link, several parameters influence
the nature of this transmission:

 Direction of Data Exchange : Determines how data flows between the transmitter and
receiver.
 Transfer Format : Specifies the number of bits transmitted together during each
operation.
 Timing Alignment : Relates to the synchronization between transmitter and receiver to
ensure data is correctly interpreted.
 Transfer Capacity : Measured in bits per second (bit/s), it indicates the amount of
information that can be transported within a given time frame.
These characteristics collectively define the conditions and performance of communication
between devices.

2.2.2.1. Parallel Transmission

Parallel transmission involves sending multiple bits simultaneously through separate channels,
with each bit traveling over its dedicated path (such as a specific wire). Cables used for this type
of transmission often consist of multiple wires grouped into flat ribbon cables. Figure 5
illustrates a diagram of parallel binary data transmission..

12
Chapter 1: State of the Art

Figure.5. Parallel transmission diagram

2.2.2.2. Serial Transmission

In a serial connection, data is transmitted sequentially, one bit at a time, through a single
communication channel. This transfer method is commonly used to establish links between
devices, such as USB (Universal Serial Bus) ports on computers, which allow various
peripherals to be connected efficiently and universally. This system is particularly suited for
communications over varying distances, offering a simple and reliable solution for data
transmission. Figure 6 illustrates a diagram of sequential binary data transmission.

Figure.6. Serial transmission diagram

2.2.3. Types of Transmission

Data can be transmitted through various communication modes, each with its own characteristics
and advantages

 Simplex Mode This is one-way communication, where information flows exclusively

from a source to a recipient without the possibility of reverse communication. This type
of connection is suitable when the data flow does not require bidirectional interaction
(e.g., radio broadcasting).

 Half-Duplex Mode This mode allows two-way communication, but not simultaneously.
In other words, only one device can transmit data at a time while the other receives (e.g.,
walkie-talkies).

 Full-Duplex Mode This system allows simultaneous two-way communication, enabling

each device to transmit and receive data at the same time. It is the type of transmission
used for interactions requiring immediate responsiveness (e.g., telephone conversations

13
Chapter 1: State of the Art

or interactions between a keyboard/mouse and a computer). Figure 7 illustrates the three

data transmission modes: Simplex, Half-Duplex, and Full-Duplex.

Figure.7. Transmission mode diagram

2.2.4. Transmission Modes
There are two primary transmission modes designed to address the challenges of data
transmission: asynchronous transmission and synchronous transmission .
2.2.4.1 Asynchronous Transmission

In this mode, data is transmitted without a shared clock reference. Each character is preceded by
a start bit (START) and followed by a stop bit (STOP) , indicating the beginning and end of the
sequence. This method allows characters to be sent irregularly, without requiring strict temporal
synchronization. Figure 8 illustrates the format of a data frame used in serial communication.

Figure.8. Asynchronous Transmission of N Bits

2.2.4.2 Synchronous Transmission

In this mode, the transmitter and receiver are synchronized to a common clock reference. This
synchronization ensures a continuous data stream without the need for additional bits to mark the
start and end of each character, thereby optimizing bandwidth utilization. Figure 9 illustrates the
data frame format used in serial communication.

Figure.9. Synchronous Transmission of N Bits

14
Chapter 1: State of the Art

2.2.4.3. Importance of Synchronization

In this type of transmission, the transmitter and receiver use a shared clock for synchronization.
This allows data to be sent continuously without adding specific bits to identify the start and end
of each character, making the process more efficient.

2.3. Application Domains

2.3.1. Digital Communication Systems

Telephone modems are used to transmit data over telephone lines, enabling digital
communication through complex modulation techniques to compensate for channel
imperfections.

2.3.2. Data Transfer in Computer Networks

Local area networks (LANs) and wide area networks (WANs) use various protocols (e.g.,
TCP/IP) to exchange data between computers and servers. This includes data transmission via
Ethernet cables or fiber optics.

2.3.3. Multimedia Transmission

Streaming platforms (e.g., Netflix or Spotify) use data transmission to deliver audio and video
content in real time, requiring high bandwidth to ensure a seamless user experience.

2.3.4. Chaotic Systems

Chaotic systems have several interesting applications in data transmission, including:

2.3.4.1. Secure Transmission Chaotic systems offer an innovative approach to securing

communications. By superimposing a chaotic signal onto the transmitted information, these
systems make data less predictable and harder to intercept, enhancing confidentiality and
integrity.

2.3.4.2. Chaotic Synchronization Chaotic synchronization methods are used to ensure that the
transmitter and receiver follow synchronized trajectories , enabling accurate recovery of
transmitted data despite disturbances and noise in the communication channel. This approach is
particularly valuable for enhancing the security of information transmissions.

2.3.4.3. Pseudo-Random Number Generators Chaotic systems can be employed to design

pseudo-random number generators (PRNGs) , which are essential for cryptographic algorithms.
These generators enhance security by producing unpredictable cryptographic keys.

15
Chapter 1: State of the Art

2.3.4.4. Chaotic Modulation This technique uses chaotic signals to modulate data before
transmission, improving robustness against interference and optimizing bandwidth utilization.

2.3.4.5. Applications in the Internet of Things (IoT) Chaotic systems can be integrated into
IoT devices to secure data transmission between connected devices, ensuring that only
authorized parties can access the exchanged information. [1]

3. Secure Communication Based on Chaotic Systems

A chaotic system is a nonlinear deterministic dynamical system whose behavior appears

random due to extreme sensitivity to initial conditions. Although governed by precise
mathematical laws, its trajectories in phase space are unpredictable over long periods and
converge toward strange attractors (bounded fractal sets). This property, combined with a
positive Lyapunov exponent, makes chaotic systems ideal for secure communication, as they
generate pseudo-random signals that are difficult to predict or intercept.

3.1. Chaotic Transmission System

A chaotic transmission system consists of a transmitter generating a chaotic signal, a

communication channel, and a receiver synchronized with the transmitter to recover the
information masked within the signal. This system relies on synchronizing two chaotic
dynamical systems (e.g., Chua or Colpitts oscillators), enabling secure data transmission thanks
to the unpredictability of the signals and their resistance to interception attacks.

3 .1.1. Chaotic Transmitters

A chaotic electronic emitter is a system that employs nonlinear circuits (oscillators, switches) to
generate chaotic electrical signals.

These circuits exploit nonlinear dynamics to produce chaotic oscillations.

Examples include Chua's oscillators, which rely on nonlinearities generated by diodes or

operational amplifiers, feedback circuits, or reactive components (inductors, capacitors).

Chaotic emitters utilize nonlinear dynamical systems to generate chaotic signals. These signals
are deterministic but appear random, making them difficult to distinguish from random noise.
The message to be transmitted is mixed with the chaotic signal. When the message's amplitude is
small relative to the chaotic signal's amplitude, the message is effectively masked.

The resulting signal is transmitted over a channel, where it appears as pseudo-random noise,
making it challenging for unauthorized parties to extract the message. The receiver employs an
identical chaos generator to synchronize with the received signal. This synchronization enables

16
Chapter 1: State of the Art

the recovery of the original message by subtracting the locally generated chaotic signal from the
received signal.[2][3]

3 .1.2. Chaotic Receivers

A chaotic electronic receiver is designed to recover information masked within chaotic signals.

To extract the message, the receiver must synchronize its behavior with that of the transmitter.
This involves generating a chaotic signal identical to the one used by the transmitter. The latter
(master) generates a chaotic signal transmitted to the receiver system (slave). The receiver
synchronizes its behavior with the transmitter to retrieve the masked message.

Several methods are employed, including synchronization via the Pecora and Carroll method,
unidirectional or bidirectional coupling, and the use of adaptive observers or sliding-mode
observers to reconstruct the complete state of the transmitter system. Once synchronization is
established, the receiver can extract the original message by subtracting the locally generated
chaotic signal from the received signal. Figure 10 illustrates a secure communication system
over a public channel using encryption and synchronization techniques.[1]

Figure.10.Secure transmission diagram using two Colpitts oscillators [4]

17
Chapter 1: State of the Art

3 .2. Basic Chaotic Systems

For data transmission, several applications are based on basic chaotic systems. This section
describes the most popular ones.

3 .2.1 Lorenz Oscillator

This system is the first chaotic system described by the following differential equations:
��
= � �−�
��
��
��
= � �−� −� (1)
��
��
= �� − ��

�ℎ����, �, ���� are the system parameters. This system generates chaotic behavior in the form
of a butterfly-shaped attractor (butterfly effect). It is widely used for meteorological simulations
and secure communications.

3 .2.2. Rössler Oscillator

This system is described by the following differential equations:

��
= −� − �
��
��
��
= � + �� (2)
��
= �+� �−�
��

with a , b , and c as system parameters. This system generates chaotic behavior with a simpler
structure than the Lorenz system. Its attractor is asymmetric.
3.2.3. Chua Oscillator
The Chua circuit is a nonlinear electronic circuit representing one of the simplest and most
famous examples of dynamical systems capable of producing chaotic behavior, introduced by
Leon O. Chua in 1983 [5].

Figure 11 illustrates the various components of the Chua circuit, including

 A resistor,
 An inductor,
 A capacitor (or two),
 A nonlinear elemen known as an "active nonlinear resistor" or "Chua's diode."

18
Chapter 1: State of the Art

Figure.11. Chua Circuit[42]

The differential equations describing the system's behavior are as follows:

� = � �−�−ℎ �
� = �−�+� (3)
� = −��
where �, �, ��� are state variables representing the voltages across the capacitors and the current
through the inductor. α , β , and γ are positive parameters related to the circuit components.

ℎ � is a nonlinear function characterizing the behavior of the active nonlinear resistor. A

classical form of e ℎ � is given by:
1
ℎ � = �1 � +
2
�0 − �1 �+1 − �−1 (4)

where �0 ���1 ​ are constants associated with the slopes of the nonlinear function [5].

4. Signature of Chaotic Systems

4.1. Dissipativity
A chaotic system is a dissipative system that loses energy over time due to irreversible processes
(e.g., friction, resistance, viscosity). In these systems, although energy is constantly dissipated,
part of it can be injected by the environment or external forces. This allows the system to
maintain complex and nontrivial dynamics, even though it loses energy globally.

In dissipative systems, the phase space gradually contracts over time due to energy losses. This
means that the system’s trajectories tend to converge toward a specific region called an attractor.
In chaotic dissipative systems, this attractor can be a strange attractor, which has a fractal
structure.

he mathematical condition for a system to be dissipative is based on the divergence of the vector
field f (t,x,u). Divergence measures how the flow of trajectories contracts or expands in phase
space. For an n -dimensional system, the divergence is given by:

19
Chapter 1: State of the Art
� ���
�⋅�= �=1 ���
(5)

where �� is the i -th component of the vector field f(t,x,u).

 If ∇⋅ f < 0 for all x and t , the system is dissipative, as the volume in phase space
contracts over time.
 If ∇⋅ f = 0, the system is conservative, meaning the volume remains constant.
 If ∇⋅ f > 0, the system is expansive, indicating that the volume increases with time.

4.2. Strange Attractor

A strange attractor is a bounded subset of phase space that attracts the system’s trajectories in the
long term while exhibiting a fractal structure and chaotic behavior. Unlike simple attractors (e.g.,
fixed points or limit cycles), a strange attractor has a non-integer (fractal) dimension and is
associated with aperiodic behavior.

In a strange attractor, although the global phase space volume is contracting (due to dissipation),
there are local fluctuations in the expansion rate. These fluctuations lead to a complex mixing of
trajectories, which is a fundamental feature of chaos.

Strange attractors spread differently in phase space and can be characterized by distinctive
features reflecting their complex, fractal nature. Examples include asymmetric attractors, spiral
attractors, and multilayer attractors. Figure 12 shows four types of strange attractors: the Lorenz
attractor, Chen attractor, Rössler attractor, and Chua attractor.

Figure.12. Asymmetric Attractors and Spiral Attractors [6]

20
Chapter 1: State of the Art

4.3. Sensitivity to Initial Conditions

All strange attractors share the property of sensitivity to initial conditions, though this
sensitivity may vary in intensity depending on the system. This means that two nearby
trajectories in phase space diverge exponentially over time, a fundamental hallmark of chaos

4.4. Lyapunov Exponents

The primary tool for quantifying chaos is the Lyapunov exponent . The Lyapunov exponent λ is
defined as the average rate of exponential divergence or convergence between two
infinitesimally close trajectories in phase space.

If λ>0, trajectories diverge exponentially, indicating chaotic behavior.

If λ<0, trajectories converge, reflecting local stability.
If λ=0, there is neither divergence nor convergence, often associated with periodic or quasi-
periodic motion.
The Lyapunov exponent is a key criterion (but not sufficient alone) for identifying chaotic
behavior.
In a three-dimensional system, there are three Lyapunov exponents, typically denoted in
descending order (�1​ ≥�2≥�3). Each exponent corresponds to an orthogonal direction in phase
space:

 λ1​ : Measures the maximum trajectory divergence. A positive �1 > 0 indicates chaos,
as nearby trajectories separate exponentially over time.
 λ2​ : May be zero or negative. A zero exponent �2 = 0 is typical for chaotic systems,
representing the direction tangent to the trajectory where no divergence or convergence
occurs.
 λ3​ : Generally negative in dissipative systems, reflecting contraction in another
direction of phase space [7]

5. Problem Statement
5.1. Research Problem

In the aeronautics industry, secure data transmission remains critical due to stringent security
requirements. This project aims to design and experiment with a basic chaotic electrical circuit,
serving as a foundational component for constructing a chaotic transmitter to ensure secure data
transmission. To achieve this, we propose studying the dynamic behavior of the simplest chaotic
circuit in secure communication: Chua's Circuit . Subsequently, this circuit will be designed and
implemented. The focus will then shift to analyzing the synchronization of two Chua circuits
21
Chapter 1: State of the Art

configured in a unidirectional master-slave mode. This synchronization is essential for

reconstructing the useful signal (masked by chaos) exclusively by an authorized receiver, thereby
enhancing communication confidentiality.

5.2. Required Tasks

The tasks include:

 Studying and analyzing the dynamic behavior of Chua’s circuit using MATLAB.

 Simulating the electrical circuit first via simulation software, then implementing it
experimentally.

 Modifying the classical system’s characteristics to enhance chaotic behavior.

6. Conclusion

The study of chaos and its signatures highlights the critical role of strange attractors, sensitivity
to initial conditions, and Lyapunov exponents in characterizing complex dynamical systems.
Chua’s system, as a representative model, provides an ideal experimental ground for
understanding the unpredictable behaviors of nonlinear systems. Its relative simplicity and
ability to generate chaotic oscillations make it a valuable tool for exploring fundamental chaos
properties. In the following chapter, we will analyze the dynamic behavior of this circuit,
focusing on the core characteristics of chaos.

22
Chapitre 2: Analyse du comportement
dynamique du système de Chua

23
Chapitre 2: Analyse du comportement dynamique du système de Chua

1. Introduction
This chapter is dedicated to the analysis of the dynamic behavior of Chua’s system, a nominal
model in chaos theory. The primary objective is to study the fundamental properties of this
nonlinear system, including its dissipativity, equilibrium points, and the complex structure of its
strange attractor. This work aims to highlight the dynamic mechanisms underlying chaos in
Chua’s system and establish the theoretical foundations necessary for its in-depth study.

2. Modeling and Stability Analysis of Chua’s System

The Chua circuit is the first electrical circuit capable of generating chaotic behavior. Several
mathematical models of this circuit have been developed. The classical configuration consists of
an inductor, two capacitors, a resistor, and a Chua’s diode. Figure 13 illustrates the components
of the Chua circuit.

Figure .13. Chua Circuit 2

2.1. Mathematical Model of the Chua System

Chua’s system is a 3-dimensional nonlinear system. Based on equations (3) and (4), the system is
described as follows:

� = � �−�−ℎ �
1
� = �−�+� et ℎ � = �1 � + 2 �0 − �1 �+1 − �−1
� = −��

where α and β are positive parameters governing the system dynamics.

The nonlinearity of the system is entirely contained within the nonlinear function h(x) , defined
as piecewise linear .

24
Chapitre 2: Analyse du comportement dynamique du système de Chua

Depending on the sign of x , the function h(x) changes.

Therefore, the following cases are obtained:

Case 1 : x > 1
1
ℎ � = �1 � + �0 − �1 � + 1 − � + 1
2 (6)
ℎ � = �1 � + �0 − �1

Case 2 : −1 ≤ x ≤ 1

1
ℎ � = �1 � + � − �1 � + 1 − −� + 1
2 0
1
ℎ � = �1 � + �0 − �1 � + 1 + � − 1
2
(7)
ℎ � = �1 � + �0 − �1 � = �0 �
Case 3 : x<−1
1
ℎ � = �1 � + �0 − �1 −� − 1 + � − 1
2
ℎ � = �1 � +
1
�0 − �1 −2 (8)
2
ℎ � = �1 � − �0 − �1 .

Based on equations (6), (7), and (8), we can conclude that h(x) is defined as follows:
1
�1 � + 2 �0 − �1 si � > 1
ℎ � = �0 � si − 1 ≤ � ≤ 1 (9)
1
�1 � − 2 �0 − �1 si � <− 1

Figure 14 represents the piecewise linear function ℎ � with �0 =− 0.714 ��� �1 = 1.143.

Figure.14. Graphical representation of a piecewise linear function ℎ �

25
Chapitre 2: Analyse du comportement dynamique du système de Chua

2.2. Dissipativité
To calculate the dissipativity of Chua’s system, we use the concept of the divergence of the
vector field governing the system. Dissipativity is determined by evaluating whether the volume
in phase space decreases over time. If the divergence is strictly negativee � ⋅ � < 0 , the system
is dissipative.

The divergence � ⋅ � measures how the flow of trajectories contracts or expands in phase space.
For a three-dimensional system, it is given by:

�� �� ��
� ⋅ � = �� + �� + �� (10)

2.2.1.Calculation of � ⋅ �​

Starting from the system’s differential equations (9), the following equations are obtained, where
h′(x) is the derivative of h(x) with respect to x :

�� �
= � �−�−ℎ � = � −1 − ℎ' �
�� ��
�� �
�⋅�= = �−�+� = −1 (11)
�� ��
�� �
= −�� = 0
�� ��

From the system’s differential equations (11), the following equations are derived, where h′(x)
remains the derivative of h(x) :

�1 si � > 1
ℎ' � = �0 si − 1 ≤ � ≤ 1 (12)
�1 si � <− 1

−� 1 + �1 si � > 1
�� ��
Thus ��​ becomes : ��
= −� 1 + �0 si − 1 ≤ � ≤ 1 (13)
−� 1 + �1 si � <− 1

By combining the results above, the expression for the total divergence is:

−� 1 + �1 − 1 si � > 1
� ⋅ � = −� 1 + �0 − 1 si − 1 ≤ � ≤ 1 (14)
−� 1 + �1 − 1 si � <− 1

2.1.3. Interpretation of Dissipativity

To compute the dissipativity, we fix the typical parameters as � = 15.6, � = 28, �0 =− 1.143
and �1 =− 0.714.

26
Chapitre 2: Analyse du comportement dynamique du système de Chua

3. Case 1 : x>1 ou x<−1

In this case, �1 =− 0.714

� ⋅ � =− � 1 + �1 − 1 =− 15.6 1 + −0.714 − 1 =− 15.6 0,286 − 1 =− 4,4616 − 1

=− 5,4616 < 0

Therefore, the system is dissipative .

4. Case 2 : −1 ≤ x ≤ 1
In this case, �0 =− 1.143

� ⋅ � =− � 1 + �0 − 1 =− 15.6 1 + −1,143 − 1 =− 15.6 −0,143 − 1 = 2,2308 − 1

= 1,2308 > 0

Therefore, the system is expansive .

3.Study of Equilibrium Points and Their Stability

To calculate the equilibrium points of Chua’s system, we solve the system of differential
equations under the condition that the time derivatives of the state variables are zero. This means
finding the values of x , y and z such that::

�=0 �=0 �=0

3.1. Determination of Equilibrium Points

We obtain the following system:

� = � �−�−ℎ � = 0
� = �−�+� = 0 (15)
� = −�� = 0

Solving this yields:

� �−�−ℎ � = 0 ⇒ � −� − ℎ � = 0 ⇒− � − ℎ � = 0 ⇒ ℎ � =− �
� − � + � = 0 ⇒ � + � = 0 ⇒ � =− � (16)
� = 0 ⇒− �� = 0 ⇒ � = 0

Thus, we must solve h(x) = − x :

1
ℎ � + � = 0 ⇒ �1 + 1 � + � − �1 �+1 − �−1
2 0

27
Chapitre 2: Analyse du comportement dynamique du système de Chua

−�0−�1
�= �1 +1
si � > 1
ℎ � + � = 0��� � �1 + 1 + �0 − �1 = 0 ⇒ � = 0 si − 1 ≤ � ≤ 1 (17)
� −�
� = �0 +11 si � <− 1
1

From equations (15), (16), and (17), we derive the following three equilibrium points:

−�0−�1 �0−�1
E1 : �, �, � = , 0, (si x >1 )
�1 +1 �1 +1
E2 : �, �, � = 0,0,0 (si - 1 ≤ x ≤ 1) (18)
�0 −�1 �0 −�1
E3 : �, �, � = �1 +1
, 0, − �1 +1
(si x < -1)

The equilibrium points of the Chua system depend on the parameter �0 ���1 which significantly
impacts the system’s dynamics.

The Point �2 corresponds to the origin of the phase space � = 0, � = 0, � = 0 , representing a

state where all variables are zero. This is a trivial equilibrium point .
In contrast, the points �1 et �3 lie off the origin, and their coordinates depend on the parameter
�0 ���1.these represent non-trivial equilibrium points.
​ �1 and �3 are symmetric with respect to the origin, reflecting the intrinsic symmetry of the
Chua system (x→−x, z→−z).

These points define the geometry of the attractor. Trajectories oscillate between them, forming
the characteristic double-scroll attractor spirals. They orbit around �1 and �3 swithout ever
converging toward them, which generates chaos.

�0 does not directly affect the position �2 , but influences its stability via the Jacobian matrix .
Meanwhile, a variation in �1 would shift these points �1 et �3 .long the x -axis
For �0 =− 1.143 Et �1 =− 0.714, we obtain the following equilibrium points:

�1: �, �, � = 1.5,0, − 1.5

�2: �, �, � = 0,0,0 (19)
�3: �, �, � = −1.5,0,1.5

3.2. Stability Analysis

3.2.1. Stability of Nonlinear Systems

The stability analysis of nonlinear systems relies on key mathematical concepts essential for
studying behavior near equilibrium points.

28
Chapitre 2: Analyse du comportement dynamique du système de Chua

A nonlinear system is generally described by ordinary differential equations (ODEs) in state-

space form:

� = � �, � (20)

where :
�
 � = �1 , �2 , …, �� is the state vector (dynamic variables of the system).
 � is a nonlinear function describing the system’s dynamics [7].

Stability analysis involves linearizing the system around its equilibrium points. Linearization is a
mathematical method used to approximate a nonlinear system locally with a linear model,
simplifying its analysis. The goal is to study the local stability of equilibrium points. This
approximation is valid only near the equilibrium point and does not capture global behavior (e.g.,
chaos) [7]

To analyze local stability around an equilibrium point � (such as � �∗ = 0 ), the system is

linearized by computing the Jacobian matrix J :

��
�=
���=�∗
(21)

This matrix approximates the nonlinear system with a linear model near the equilibrium
point �∗ : �� ≈ � ⋅ �� (22)

where �� = � − �∗ represents a small perturbation around the equilibrium [8].

The eigenvalues �� of the Jacobian matrix determine the local stability of the equilibrium point:

 Asymptotic Stability Case : If at least one eigenvalue has a positive real part ℜ �� < 0,
the point is stable.

 Instability Case : Si au moins une valeur propre a une partie réelle positive ℜ �� > 0
the point is unstable.

 Critical Case : If eigenvalues have zero real parts ℜ �� = 0 , stability cannot be

determined by linearization alone (nonlinear methods, such as Lyapunov’s direct method,
are required).

3.2.2 Stability Analysis Around Each Equilibrium Point

The Jacobian matrix J is constructed from the partial derivatives of the system’s equations with
respect to the state variables x , y , and z . Based on equation (3), we have:

29
Chapitre 2: Analyse du comportement dynamique du système de Chua

�� �� ��
�� �� ��
�� �� ��
�= �� �� ��
(23)
�� �� ��
�� �� ��

�=
�� � �� � ��
= � � �−�−ℎ � =− � 1 + ℎ' � = � � �−�−ℎ � =� =0
�� � �� � ��
�� � �� � �� �
= � �−�+� =1 = � � − � + � =− 1 = � �−�+� =1
�� � �� � �� �
�� �� � ��
=0 = � −�� =− � =0
�� �� � ��
The final Jacobian matrix is:

−� 1 + ℎ' � � 0
�= 1 −1 1 (24)
0 −� 0

The Jacobian matrix � xplicitly depends on the parameters ����. These parameters modulate the
eigenvalues λ , which are solutions to the characteristic equation: ��� � − �� = 0.

A variation in � �� � modifies the coefficients of this equation, altering the eigenvalues and
therefore the system’s stability.

 For the equilibrium point �1 = 1.5,0, − 1.5 ,the following Jacobian matrix is
obtained:

−15.6 1 + −0.714 15.6 0 −4.4616 15.6 0

�1 = 1 −1 1 = 1 −1 1
0 −28 0 0 −28 0

Thus

4.4616 − � 15.6 0
det �1 − �� = det 1 −1 − � −28 = 0
0 −� −�

��� �1 − �� = −4.4616 − � �2 + � + 28 + 15.6�

− �3 − 5.4616�2 − 16.8616� − 124.9248 = 0

The solutions of the characteristic equation are : �1 ≈− 6.00, �2 ≈ 0.3 + 4.52�, �3 ≈ 0.3 −
4.52�

 For the equilibrium point �2 = 0,0,0 ,the following Jacobian matrix is obtained:

30
Chapitre 2: Analyse du comportement dynamique du système de Chua

−� 1 + −1.143 � 0 −15.6 ⋅ 1 − 1.143 15.6 0

�2 = 1 −1 1 = 1 −1 1
0 −� 0 0 −28 0
2.2308 15.6 0
= 1 −1 1
0 −28 0

2.2308 − � 15.6 0
det �2 − �� = det 1 −1 − � −28 = 0
0 −� −�

det �2 − �� = 2.2308�2 + 2.2308� + 62.4624 − �3 + �2 + 28� + 15.6�

− �3 + 1.2308�2 − 10.1692� + 62.4624 = 0

The solutions of the characteristic equation are : �1 ≈ 3,47, �2 ≈ 1.12 + 4.08�, �3 ≈ 1.12 −
4.08�

 For the equilibrium point �3 = −1.5,0,1.5 ,the following Jacobian matrix is obtained:

4.4616 15.6 0
�3 = 1 −1 1 = �1
0 −28 0

We obtained the same Jacobian matrix for symmetric points �1 = �3 , therefore �1 ��� �3 share
the same eigenvalues �1 =�3 .

The stability study is performed around each equilibrium point as follows:

 For points �1���3: The presence of an eigenvalue with a negative real part (� �1 ≈−
6 < 0 ) results in a stable direction around the equilibrium point. However, the two
complex conjugate eigenvalues �2 ���3 have a positive real part, which means they
represent unstable directions with chaotic dynamic behavior. Although these points have
a stable direction, (� �1 > 0) , the unstable directions dominate (� �2 ���(�3 ) < 0) .
rajectories around this point are temporarily attracted before diverging.

 For points �2: The presence of an eigenvalue with a positive real part � �1 ≈ 3,47 > 0
creates an unstable direction around the equilibrium point. The two complex conjugate
eigenvalues �2 ���3 also have positive real parts, indicating unstable directions associated
with chaotic dynamic behavior.

31
Chapitre 2: Analyse du comportement dynamique du système de Chua

Therefore, all points �1, �2���3 are unstable due to eigenvalues with positive real parts. This
instability is the driving force behind chaos in Chua’s system, where trajectories oscillate
between the scrolls of the double-scroll attractor without ever converging to an equilibrium.

4. The Chua Strange Attractor

4.1. MATLAB-Matcont

To simulate the double-scroll attractor of Chua’s system, we employed MATLAB-Matcont.

Matcont is a graphical MATLAB toolbox designed for interactive numerical analysis of
nonlinear dynamical systems. This package enables numerical solution of the system’s
differential equations, visualization of trajectories in phase space, and analysis of chaotic
properties. The prediction-correction continuation algorithm employed by MATCONT facilitates
precise computation of trajectories and critical points, allowing in-depth analysis of strange
attractors without requiring specialized programming expertise. This interactive approach
significantly simplifies the study of chaotic systems and provides a flexible platform to explore
and understand their complex behavior.

4.2. Double-Scroll Strange Attractor of Chua via Matcont

To generate Chua’s double-scroll attractor, the parameter values were set as 15.6, � = 28, �0 =
− 1,1 , ��� �₁ =− 0,714 ,to ensure chaotic dynamics .

In accordance with the conditions required for chaos. Initial conditions for the first simulation
were chosen as (x₀, y₀, z₀) = (0.1, 0, 0). These parameters and conditions yield a characteristic
strange attractor illustrated in Figure 15 (espace des phases x-y-z ). The attractor exhibits a
symmetric fractal structure relative to the origin, with two main scrolls centered around the
equilibrium points (±1.5, 0, 1.5), rconnected by oscillating trajectories. Figures 16, 17, and 18
depict the phase portraits in the (x-y ),( y-z) and (z-x ), respectively, highlighting the attractor’s
geometric complexity and symmetry.

32
Chapitre 2: Analyse du comportement dynamique du système de Chua

Figure.15.Double-scroll attractor in phase space x-y-z

Figure.16. Double-scroll attractor x-y

Figure.17. Double-scroll attractor y-z

33
Chapitre 2: Analyse du comportement dynamique du système de Chua

Figure.18. Double-scroll attractorl z-x

 Shape and Structure

The double-scroll attractor resembles intertwined spirals, forming a complex fractal

structure in the three-dimensional phase space (x,y,z).These "scrolls" are separated by a
central region where trajectories rapidly oscillate between the two lobes before
converging toward one spiral.

 Position of Equilibrium Points in the Attractor

The equilibrium points �1����3 lie at the centers of the two scrolls. Trajectories spiral
around these points, forming orbits that never converge to them (due to instability).
Figure 14 illustrates that the scrolls are centered around these points �1���3 ​ , with
oscillating trajectories connecting the two. The origin �2 acts as a pivot between the
scrolls. Trajectories pass near �2​ , but do not stabilize there (instability persists).

 Fractal Nature

he attractor is termed "strange" because it exhibits a non-integer fractal dimension. This

reflects its geometric complexity: though confined to a bounded region of phase space, it
is neither a simple surface nor a line. While the global phase space contracts (as the
system is dissipative), local regions allow temporary stretching or expansion. These
fluctuations contribute to the intricate mixing of trajectories, a hallmark of chaos.

34
Chapitre 2: Analyse du comportement dynamique du système de Chua

 Symmetry

The double-scroll attractor is symmetric relative to the origin. This symmetry arises from
Chua’s system’s intrinsic property : (x,y,z)→(-x,-y,-z). If a trajectory passes through a
point (−x,−y,−z).

 Complex Oscillations

Trajectories on the double-scroll attractor oscillate rapidly between the two lobes,
creating rich chaotic dynamics. These oscillations are neither periodic nor quasi-periodic
but remain confined within the attractor’s bounds.

 Rapid Transitions

Transitions between the two lobes are rapid and unpredictable, further enhancing the
system’s complexity. These transitions are driven by the nonlinear function ℎ � , which
introduces local discontinuities in the vector field.

4.3. Sensitivity to Initial Conditions

The double-scroll attractor is typically chaotic, meaning that two trajectories starting from very
close initial conditions will eventually diverge exponentially over time. This behavior is
confirmed by the presence of a positive Lyapunov exponent �1 > 0 .

As an example, consider the following two sets of initial conditions:

� = 0,1 � = 0,1 + 10−14

��1 � = 0 ��2 � = 0
� = 0 � = 0

Figures 19, 20, and 21 illustrate the temporal evolution of the state variables (x, y, ou z)for these
two slightly different initial conditions. The divergence of trajectories becomes evident over time,
highlighting the system’s extreme sensitivity to initial conditions a defining feature of
deterministic chaos in Chua’s system.

35
Chapitre 2: Analyse du comportement dynamique du système de Chua

Figure.19.Sensitivity to initial conditions for variable x (blue curve corresponds to IC1 and red

curve to IC2)

Figure.20. Sensitivity to initial conditions for variable y (blue curve corresponds to IC1 and
green curve to IC2)

Figure.21. Sensitivity to initial conditions for variable x (blue curve corresponds to IC1 and red
curve to IC2)

36
Chapitre 2: Analyse du comportement dynamique du système de Chua

For the three figures, the two trajectories start very close together at the beginning (initial time),
reflecting similar initial conditions. However, after a certain time ( � = 45�), .a progressive shift
emerges between the two curves. Over time, the differences between the two trajectories increase
rapidly, illustrating sensitivity to initial conditions a defining characteristic of chaos.

The nonlinearity of ℎ � plays a crucial role in amplifying small local fluctuations. The abrupt
transitions between regions � > 1, − 1 ≤ � ≤ 1, � <− 1 create "jumps" in the vector field.
These jumps amplify small perturbations in trajectories, making the system extremely sensitive
to variations.

The two curves exhibit complex and irregular oscillations, typical of strange attractors. Although
the trajectories remain confined within the double-scroll attractor region, they follow no apparent
periodicity or quasi-periodicity.

4.4. Lyapunov Exponents

For a dynamical system, imagine two points initially very close to each other. Lyapunov
exponents measure how quickly the distance between these two points evolves over time.

If the largest Lyapunov exponent is positive ( �max > 0 ), a small perturbation in initial
conditions will grow exponentially.

To compute the Lyapunov exponents, the Wolf algorithm was employed. This calculation
depends on the system’s differential equations, attractor parameters � = 15.6, � = 28, �0 =−
1.143, �1 =− 0.714 , and initial conditions. The algorithm implemented in MATLAB is
detailed in Appendix 1 .

These values are obtained after prolonged numerical simulation and reflect the following
characteristics: �1 ≈ 0,433, �2 ≈ 0,0435, �3 ≈− 4,267.

 �1 > 0: Indicates the system is chaotic, as two nearby trajectories diverge exponentially.

 �2 ≈ 0 :Corresponds to the direction tangent to the trajectory (a feature of continuous

systems).

 �3 < 0 : Shows global contraction in another direction, confirming the system’s

dissipativity.

37
Chapitre 2: Analyse du comportement dynamique du système de Chua

4.5. Kaplan-Yorke Dimension

The fractal dimension of the attractor can be estimated using the Lyapunov exponents. For
instance, the Kaplan-Yorke dimension is defined as:
�
�=1 ��
�� = � + ��+1
(25)

�
where j is the largest integer such that �
�=1 �
≥0

�
We seek the largest integer j satisfying �
�=1 �
≥0:

1 2
For j=1, j=2 and j=3 �
�=1 �
= �1 = 0.4333 ≥ 0 �
�=1 �
= �1 + �2 = 0.4333 +
0.0435 = 0.4768 ≥ 0
3

�� = �1 + �2 + �3 = 0.4333 + 0.0435 − 4,267 =− 3,7902 < 0

�=1
�
Here, j=2 is the largest integer fulfilling �
�=1 �
≥0

Application of the Formula

Substituting � = 2, �1 = 0.4333, �2 = 0, ∧ �3 = 0.0436

� 2
�
�=1 �
�
�=1 � 0.4769
��� = � + =2+ =2+ = 2 + 0.1117 ≈ 2.11
��+1 �3 4,2680

In some studies, this value is approximated as 2.3, depending on numerical methods used to
estimate the Lyapunov exponents.

The Kaplan-Yorke dimensione ��� ≈ 2.3 indicates that the double-scroll attractor occupies an
intermediate region between a surface D = 2 and a full volumet D = 3. This fractal characteristic
is a key feature of strange attractors, reflecting their complex, non-integer geometric structure [9].

5. Richness of Dynamic Behavior Beyond Chaos

Chua’s system is not limited to chaotic regimes. By varying parameters such as �, �, �0 ���1 ​ ,
the system can exhibit a wide range of dynamic behaviors, transitioning from simple stability to
chaos, including limit cycles and quasi-periodic regimes. This diversity highlights the inherent
complexity of nonlinear systems.

38
Chapitre 2: Analyse du comportement dynamique du système de Chua

5.1. Fixed Point

For low values of � ��. � = 5 and � ��. � = 10 , with �0 =− 0,5 and �1 =− 1,5 , the
system converges to a stable fixed point,typically located at the origin �2​ = 0,0,0 .
Eigenvalue analysis reveals that the Jacobian matrix evaluated at �2 has three eigenvalues with
negative real parts and � ⋅ � < 0 , ensuring that �2 is a stable fixed point. This regime
corresponds to a state where dissipative forces dominate entirely, suppressing all oscillations or
chaotic behavior. Figure 22 illustrates the convergence toward �2 = 0,0,0 .

Figure.22. Stable dynamic

5.2. Cycles Limites et Oscillations Périodiques

A small change in �0 suppresses chaos and generates periodic oscillations , also known as
limit cycles .
LThe competition between dissipation and nonlinearity creates complex but predictable
oscillations.
If �0 =− 1.05 ��� �1 =− 0.714 with � = 15,6��� = 28,
Dissipation remains significant, but the nonlinearity h(x) introduces self-sustained oscillations ,
creating closed cycles in phase space.
Trajectories form patterns that converge toward a closed orbit.
Figures 23 and 24 show the trajectory behavior as a limit cycle in 3D and periodic oscillations
in the 2D (x, y) plane .

39
Chapitre 2: Analyse du comportement dynamique du système de Chua

Figures.23. 3D limit cycles Figures.24. 2D periodic oscillation (x,y)

6. Conclusion

The dynamic analysis of Chua’s system focuses on the key characteristics of chaos. First, the
calculation of divergence shows that the system is dissipative, which explains its convergence
toward a strange attractor. This attractor perfectly illustrates the system’s fractal nature, with a
non-integer Kaplan-Yorke dimension, extreme sensitivity to initial conditions, and positive
Lyapunov exponents. These results highlight the intrinsic complexity of Chua’s system, where
simple nonlinear interactions give rise to rich and unpredictable dynamics.

In Chapter 3, the implementation of the Chua circuit is presented, based on the corresponding
differential equations.

40
Chapitre3: Réalisation du circuit du chua et Synchronisation

Chapitre3: Réalisation du circuit du chua et

Synchronisation

41
Chapitre3: Réalisation du circuit du chua et Synchronisation

1. Introduction

This chapter presents a comprehensive study of Chua’s circuit, divided into several
complementary sections. It begins with a description of the simulation environment used
(Multisim), followed by a detailed analysis of Kennedy’s diode (used to implement Chua’s diode)
and the gyrator, which replaces the traditional inductor. Next, the complete Chua’s circuit is
presented in its entirety, along with an interpretation of the simulation results—particularly the
observation of a double-scroll attractor. The final section focuses on the synchronization of two
Chua’s circuits configured in a unidirectional master-slave setup. This synchronization is
achieved using a voltage follower amplifier , allowing the slave circuit to faithfully reproduce the
chaotic behavior of the master circuit..

2. Multisim

NI Multisim is an electronic schematic capture and simulation software package that is part of
a suite of circuit design tools, including NI Ultiboard. It provides an intuitive interface for
assembling components (resistors, capacitors, operational amplifiers, etc.) and for observing
signals at various nodes of the circuit in real time.

In the context of this work, Multisim was used to simulate Chua’s circuit in order to validate the
theoretical behavior and explore its chaotic dynamics, with particular attention given to the
emergence of the double-scroll attractor. Thanks to its advanced simulation capabilities,
Multisim provides a realistic virtual testing environment to assess the feasibility and
effectiveness of our design before physical implementation.

3. Design of the Chua Circuit

Figure.25. Simplified Chua’s circuit [10]

42
Chapitre3: Réalisation du circuit du chua et Synchronisation

The Chua circuit consists of the following components:

 Two capacitors s �1 , �2
 A gyrator , which emulates an inductore�
 A linear resistor �
 A nonlinear resistor , known as the "Chua’s diode", modeled by a piecewise-linear
function � ��

Modeling the Circuit Using Differential Equations

Starting from the differential equations describing Chua’s circuit (equation 3), we derive the
system of differential equations that models the behavior of this electronic circuit. This will be
done using fundamental electrical laws.

Definition of Nodes and Variables,We define the following state variables:

 Voltage �1 ​ at point A : voltage across capacitor �1

 Voltage �1 ​ at point B : voltage across capacitor �2
 �� ​ : current flowing through the inductor �

Applying Kirchhoff's Current Law (KCL) at node A, where the sum of incoming and outgoing
currents is zero,By choosing the convention where incoming currents are positive, we write:
��1 �2−�1
�1
��
=
�
− � �1 (26)

Applying KCL at node B gives:

��2 �1−�2
�2
��
=
�
+ �� (27)

Equation for the Inductance �

For the loop that includes the inductance, we apply Kirchhoff’s voltage law. The voltage across
the inductor is related to the variation of the current by the following relation:
���
� ��
=− �2 (28)

Final System of Differential Equations

By combining the equations obtained at the nodes and for the inductance, we obtain the
following system:

43
Chapitre3: Réalisation du circuit du chua et Synchronisation
��1 �2 −�1
�1 = − � �1
�� �
�� �1 −�2
�2 2
��
=
�
+ �� (29)
���
� =− �2
��

And by analogy:

�2�2 �2
�=
�
,� =
�1
(30)

To implement the time derivatives present in Chua's equations, capacitors combined with
operational amplifiers in integrator configuration are used: each of the two capacitors �1 et
�2

« integrates» the incoming current to produce a varying voltage x or y , in accordance with

Kirchhoff’s laws.

The Chua’s diode , built from a network of resistors and operational amplifiers, introduces the
essential nonlinearity � �1 required for chaos. Its piecewise-linear behavior with different slope
segments replicates the function � qui, which, together with the capacitors, generates the
characteristic double-scroll attractor.

Finally, to simulate the inductor L an ideal coil without resistance �� , a gyrator circuit is used
(composed of a capacitor C₃ and two operational amplifiers): this subcircuit behaves like an
active inductor, converting voltage �2 into current �1 .

4. Description of the Different Components of the Circuit

4.1. Kennedy Diode

In the study of Chua’s circuit, the nonlinearity essential to chaotic behavior is provided by a
component whose characteristic resembles a piecewise-linear negative resistance. In other words,
a device that reproduces the expected current-voltage curve is referred to as a Chua’s diode .
Several methods exist to implement such an element; in this project, we adopt the solution
proposed by M. P. Kennedy in 1992 [11].

This approach, known as the Kennedy diode , relies solely on resistors and operational amplifiers.
It involves connecting two subcircuits in parallel, each generating a segment of negative
resistance, as illustrated in Figures 26 and 27. The resulting structure emulates the piecewise-
linear i−v (current-voltage) characteristic required for chaos generation.

The equation describing the Kennedy diode, which relates voltage to current, is given as follows:

�� � = � ∗ ����ℎ� = � = 1 �6 (31)

44
Chapitre3: Réalisation du circuit du chua et Synchronisation

Figure.26. Assembly of a negative resistance

Figure.27. Characteristic i=f(U) of a negative resistance

45
Chapitre3: Réalisation du circuit du chua et Synchronisation

Figure.28. Kennedy diode

Figure.29. Characteristic i=f(U) of a Kennedy diode

46
Chapitre3: Réalisation du circuit du chua et Synchronisation

4.2. Chua’s Diode

The superposition of these two parallel branches allows for a global characteristic that meets the
requirements of Chua’s diode, as shown in Figures 28 and 29. This configuration reproduces the
piecewise-linear nonlinearity essential for the circuit’s chaotic behavior.

According to Equation (4), the equation describing Chua’s diode is given as follows:
1
ℎ � = �1 � + 2 �0 − �1 �+1 − �−1

After applying fundamental laws of electricity, we conclude that :

−�2 �5 −�2 1 �3 �6
�1 = � − , �0 = � + , ��1 = ���� , ��2 = ���� , (32)
1 �3 �4 �6 1 �3 �4 �2 +�3 �5+�6

4.3. Gyrator

Instead of using physical inductors with low internal resistance, the inductor in Chua’s circuit
can be replaced by a gyrator . This active circuit, made up of operational amplifiers and passive
components, emulates the behavior of an ideal (lossless) inductor while avoiding the parasitic
resistances associated with real coils.

�� �
The operation of the gyrator is based on the fundamental inductor equation:� � = � ��
(35)

where L is the inductance. Unlike a real coil, whose behavior depends on the magnetic field, the
gyrator we have chosen combines an operational amplifier with four resistors �7, �8, �9, �2�
et un condensateur �3. and ,which presents an impedance ��3 such that:
1
��3 = ��� (33)
3

On appliquant la loi des nœuds , l’impédance d’entrée se montre égale à

�2 �2 �8 �7�9 �2� �3 �
��� �� = ������ �� = �
�1 1
=� ��� = � (34)
7 �9 �2� ��3 � �8

In other words, the simple capacitor C behaves exactly like an inductor with an equivalent
inductance value of:
�7 �9 �2� �3
�=
�8

where �9​ , �7​ , �8​ , �9​ , ∧ �2� are passive resistors arranged in a network (as shown in
Figure 25).

47
Chapitre3: Réalisation du circuit du chua et Synchronisation

Figure.30. Antoniou gyrator implementation to simulate an ideal inductance

5.Global Chua Circuit

Together, these three blocks (two integrators, the nonlinear network, and the gyrator) implement
the Chua differential system in analog form and allow observation of its double-scroll chaotic
attractor.

Based on the previous differential equations, we can conclude that

−�2 �5 −�2 1 �3 �6 �2 �2 �
�1 = − , �0 = � +� , ��1 = ���� , ��2 = ���� , �= , � = �2
�1 �3 �4 �6 1 �3 4 �2 +�3 �5+�6 � 1

48
Chapitre3: Réalisation du circuit du chua et Synchronisation

Figur.31. Assembly of Chua’s circuit on Multisim

6. Observations of Simulation Results on Multisim

Simulations performed in Multisim reveal a progressive transition of the Chua circuit between
different dynamic regimes depending on the resistance R. At R = 200 Ω, the figures show a
stable limit cycle (Figure 32), characterized by periodic oscillations, confirming the local
stability of the equilibrium points. For R = 191 Ω and 190 Ω (Figure 33), the emergence of a
Rössler attractor indicates a quasi-periodic dynamic, marking the approach of the chaotic
threshold. When R drops below 190 Ω (Figures 34-35), the typical double-scroll attractor of
chaos appears, illustrating increased sensitivity to initial conditions and a fractal structure.
Finally, transient phenomena and hysteresis loops are observed at R = 175 Ω (Figure 37).

49
Chapitre3: Réalisation du circuit du chua et Synchronisation

Figure.32. Limit cycle (periodic regime) for R=200 �

(a) (b)

Figure.33. Rössler attractor (a) for R=190 � and (b) for R=191�

(a) (b)
Figure.34. Double-scroll attractor (a) forr R= 189 � and (b) for R=170 �

50
Chapitre3: Réalisation du circuit du chua et Synchronisation

(a) (b)
Figure.35. Double-scroll attractor (a) for R= 175 � and (b) for R= 155 �

(a) (b)

Figure.36. (a) Main (b) Delayed for R= 175 �

(a) (b)

Figure.37. Hysteresis (a) for R= 175 � and linear line (b) for R= 155 �

51
Chapitre3: Réalisation du circuit du chua et Synchronisation

7. Experimental Validation of the Chua Circuit

In the electronics laboratory at ESAT, I successfully assembled the Chua circuit by rigorously
following the theoretical schematic and selecting appropriate components. The observations
obtained on the oscilloscope, presented below in the form of figures, highlight both the limit
cycle, the characteristic chaotic attractor, the hysteresis loop, and the linear segment of the
nonlinear component. These experimental results validate both the design of the setup and the
accuracy of the mathematical model in reflecting the various dynamic behaviors observed.

Figure.38. Assembly of Chua’s circuit

52
Chapitre3: Réalisation du circuit du chua et Synchronisation

(a) (b)

Figure.39. (a) et (b) Limit cycle (periodic regime)

Figure.40. Rössler attractor

(a) (b)
Figure.41. (a) and (b) Main double-scroll attractor

53
Chapitre3: Réalisation du circuit du chua et Synchronisation

(a) (b)

(c) (d)
Figure.42. (a) , (b) , (c) and (b) Double-scroll attractor

(a) (b)

Figure.43. (a) Hysteresis and (b) Linear line

54
Chapitre3: Réalisation du circuit du chua et Synchronisation

8.Synchronization

The synchronization of chaotic systems involves coupling two (or more) nonlinear oscillators so
that they adjust their phase trajectories, despite the extreme sensitivity to initial conditions that
characterizes chaos. It has applications in cryptography, signal processing, and the modeling of
natural phenomena, where one seeks to exploit or control the complex behavior of chaotic
circuits. Several coupling schemes can be distinguished: unidirectional (master–slave),
bidirectional (reciprocal), and with state feedback (closed-loop).

8.1. Complete Synchronization

Complete synchronization between two dynamic systems is achieved by applying a

unidirectional coupling. This is the most basic and commonly encountered form of
synchronization when analyzing two systems.

Le schéma synchronisation complète a été introduit par Pecora et Carroll en 1990, qui ont
démontré la possibilité de synchroniser des oscillateurs chaotiques de Chua ou de Rössler en ne
couplant qu’une seule équation de l’état. L’unidirectionnel est particulièrement adapté aux
applications de communication sécurisée, car l’esclave ne renvoie rien au maître, ce qui empêche
toute récursion non désirée. [12]

8.2. Synchronisation unidirectionnelle maître–esclave

DThe complete synchronization scheme was introduced by Pecora and Carroll in 1990, who
demonstrated the possibility of synchronizing chaotic oscillators such as Chua’s or Rössler’s by
coupling only a single state equation. The unidirectional scheme is particularly well-suited to
secure communication applications, as the slave sends no signal back to the master, thus
preventing any unwanted recursion. [12]

If the conditions for the stability of the error � = �slave − �master are satisfied, then ��� � � =0
�→∞

Here is an explanatory diagram of bidirectional synchronization.

Figure.44. Unidirectional synchronization [3]

55
Chapitre3: Réalisation du circuit du chua et Synchronisation

8.3. Bidirectional Synchronization

In bidirectional coupling, each oscillator sends information to the other , it’s a reciprocal
coupling (mutual feedback). Each system acts as both master and slave at the same time. Two
main modes can be distinguished:

 Symmetric diffusive coupling: both systems linearly share the difference of their states
�1 = � �1 + � �2 − �1 and do so symmetrically for each direction. �2

 Single-component coupling: only one variable (e.g., the voltage) is coupled in both
directions.

Bidirectional synchronization increases the robustness of chaotic systems against disturbances

and parameter variations. It also promotes the emergence of cluster synchronization phenomena,
where subgroups of systems within a chaotic network manage to synchronize with each other.
This type of coupling is essential in large networks, as it helps structure the overall chaotic
behavior into synchronized subgroups, thus ensuring resilience and consistency in distributed
secure communications. [13]

Figure.45. Bidirectional synchronization [3]

8.4. Other Types of Synchronization

Phase synchronization: only the phases �masteret �slave remain locked, even if the amplitudes
differ. This is used in biomedical signal processing and in communications.

Impulse synchronization: the coupling occurs only at discrete moments (pulses), which reduces
the required bandwidth in secure communications.

Generalized synchronization: there exists a nonlinear function � such that �slave = � �master
instead of a direct equality.

56
Chapitre3: Réalisation du circuit du chua et Synchronisation

8.5. Implementation of Unidirectional Master–Slave Synchronization

In this scheme, a single state variable from the master (for example, x₁) is directly injected into
the same equation of the slave without any explicit feedback loop. The coupling term is typically
written as:

�slave = � �slave + � �1master � − �1slave � �1 (35)

where �2 is the selection vector of the first component.

The stability of the synchronization error then depends on the sign of the Lyapunov exponents
associated with the uncoupled sub-dynamics. This minimal coupling is simple to implement in
analog form (a summing amplifier + a gain), but may require fine tuning of the gain k to ensure
convergence. It is particularly well-suited to early prototypes of complete synchronization but
remains sensitive to parametric uncertainties if the choice of the coupled variable is suboptimal.

Figure.46. Synchronization via follower circuit for Chua’s circuit

9.Implementation of Follower Coupling on the Chua Circuit

In this section, we will detail the implementation of follower coupling synchronization applied to
the Chua circuit. Networking two chaotic circuits generates a new dynamic system with five or
six variables: its equations retain the structure of the original circuit while introducing slight
modifications. The challenge is therefore to design a setup in which the equations of the master
and the slave become strictly identical.

We chose master–slave synchronization for its simplicity: the master circuit evolves completely
autonomously, while the slave replaces its own variable � (voltage �1 ) with that of the master
via a voltage follower amplifier. This forced injection of � naturally causes the other variables �
and � in both subsystems to converge. As demonstrated in Section 3, as long as all the
parameters of both circuits are strictly identical, the voltage across �2 eand the current through
the inductor will coincide (within a negligible margin of error), thereby validating the
effectiveness of this synchronization scheme.Here is a diagram of a Chua circuit synchronization
via follower coupling.

57
Chapitre3: Réalisation du circuit du chua et Synchronisation

Figure.47. Follower coupling of Chua’s circuit [11]

Figure.48. Synchronization circuit for Chua

After building the first prototype of the Chua circuit, we assembled a synchronized system
according to the setup illustrated in Figure 48. When the components of the master and slave are
identical, synchronization occurs automatically: the voltage �2' of the slave circuit faithfully
reproduces the evolution of the master’s voltage. To achieve this, we used the same values for
the capacitance, inductance, and potentiometer to minimize the difference between the two
voltages. �2 and �2 '​ .

The results are shown in Figure 35 for R=1,75 � :two voltage curves as functions of time are
visible, along with a superimposed curve of the two signals showing a 0.34 V offset to
demonstrate synchronization between the two signals, as well as the strange attractor of both
circuits. The figure also presents cases of no synchronization between master and slave and
synchronization achieved by opening and closing a switch. The first circuit is shown in orange
and the second in red.

58
Chapitre3: Réalisation du circuit du chua et Synchronisation

(a) (b)

Figure.49. (a) Curve �2 = � � for the first circuit and (b) �2 = � � for the second circuit

(a) (b)
Figure.50. (a) et (b) Double-scroll attractor of the first and second circuits

Figure.51. Superposition curve of two signals (Synchronization)

59
Chapitre3: Réalisation du circuit du chua et Synchronisation

(a) (b)

Figure.52. (a) Case of no synchronization and (b) case of synchronization

At the start of the simulation, with the switch left open, oscilloscopes XSC2 and XSC3
respectively display the double-scroll attractors of the master and slave circuits. The last figures
then illustrate the impact of the switch state on synchronization: in the open position, no locking
occurs, whereas in the closed position, a perfect overlap of the signals is observed. For this,
oscilloscope XSC1 is connected simultaneously across capacitor C4 of the master and capacitor
C1 of the slave.

10. Conclusion

This chapter has demonstrated, from modeling to experimentation, the complete realization of
the Chua circuit and the control of its double-scroll chaotic dynamics. Thanks to Multisim and
the use of the Kennedy diode and a gyrator, we validated both the differential equations and the
system’s temporal behavior. The implementation of master–slave synchronization schemes,
whether by direct follower or unit feedback loop, confirmed that a careful injection of V₁ suffices
to lock the trajectories of V₂ and the inductive current when the parameters are identical. The
experimental results, supported by the analysis of attractors and superimposed signals, establish
the robustness of these methods for applications in cryptography or chaotic signal processing.

60
General Conclusion
This final year project has enabled the consolidation of the theoretical and practical foundations
necessary for the use of chaotic systems in secure communications, with a particular focus on the
Chua circuit. The in-depth study of its dynamic properties revealed its capacity to generate stable
chaotic behavior, characterized by a double-scroll strange attractor, controlled dissipativity, and
positive Lyapunov exponents. These results, validated through simulations in MATLAB-Matcont
and practical experiments on Multisim, highlight the relevance of the Chua circuit as a tool for
data encryption.

The hardware realization of the circuit, thanks to the use of common electronic components (a
gyrator to simulate the inductance, Kennedy diode for nonlinearity), confirmed its feasibility
under real conditions. Tests conducted across different dynamic regimes (fixed point, limit cycle,
chaos) demonstrated the system’s sensitivity to parameter variations, notably the resistance R, a
key criterion for adjusting the level of chaos. In particular, the transition to chaotic regime was
observed for R values below 190 Ω, with the formation of a symmetric double-scroll attractor
around the unstable equilibrium points ±x.

Moreover, the implementation of a unidirectional master–slave synchronization scheme

demonstrated the effectiveness of masked signal recovery. The use of a follower amplifier to
inject the master circuit’s state variable x into the slave allowed for fast and robust
synchronization, even in the presence of slight parametric variations. The experimental results,
illustrated by the superposition of time-domain signals and the similarity of attractors, validated
the receiver’s ability to faithfully reconstruct the original message. This synchronization, based
on the theory of Pecora and Carroll, relies on the exponential convergence of the synchronization
error, guaranteed by the stability of negative Lyapunov exponents for the uncoupled variables.

Chaotic signals could also be used to generate secure cryptographic keys. This research thus
validates the potential of chaotic systems for future cybersecurity, combining rigorous
mathematical theory with practical engineering.

61
Bibliography
[1] LAURENT OUDRÉ, Synchronization of chaotic systems by observers and applications to
information transmission, thesis.hal, 09/02/2013.
[2] LAURENT LARGER, Chaotic dynamics applied to cryptography, University Marie and
Louis Pasteur, 2025.
[3] AIT HAMMI ABDELFATEH, Study and implementation of a chaotic system based on the
Chua circuit, University Mouloud Mammeri, 2013–2014.
[4] OUERDIA MEGERBI, Secure data transmission based on two Colpitts chaotic systems,
Springer Nature link, Jan. 2013.
[5] ALEXANDRE SORDI, Chua oscillator: An introduction to chaos theory, SciELO, 2021.
[6] MERIEM MELIOUH, Control of the Chua chaotic system, University Mohamed Khider,
Biskra, September 2020.
[7] K. M. LIEW and XIAN-BIN LIU, Maximum Lyapunov exponent for a three-dimensional
stochastic system, ResearchGate, September 2004.
[8] WARREN WECKESSER, Math 312 Course Notes: Linearization, Department of
Mathematics, Colgate University, March 23, 2005.
[9] STEVEN H. STROGATZ, Nonlinear dynamics and chaos, Taylor & Francis Group, 2015.
[10] V. SIDERSKIY, Construction of the Chua circuit, chuacircuits, 2025.
[11] JUAN JOSÉ GUILLEN GARCÍA, Chua circuit: Study of synchronization of chaotic signals
in electrical circuits, Universidad Politécnica de Madrid, July 2022.
[12] AMROUNE NAHLA and LAINOUS MANA, Hybrid synchronization of chaotic systems,
Abd elhafid Boussouf University Center, Mila, 2022/2023.
[13] YONGGUANG YU, Synchronization of linearly bidirectionally coupled chaotic systems,
ScienceDirect, Vol. 1, October 2004.

62
Appendix
Algorithm for calculating Lyapunov exponents:
%% System parameters
m0 = -1.143; % Chua parameter
m1 = -0.714; % Chua parameter
alp = 15.6; % alpha parameter
be = 28; % beta parameter

%% Numerical parameters
dt = 0.01; % time step
T_total = 100; % total integration time (adjust as needed)
N = round(T_total/dt); % number of iterations
T_R = 0.1; % re-orthonormalization period (in time units)
steps_per_orth = round(T_R/dt); % iterations between re-orthonormalizations

%% Initial conditions
x0 = 0.1; y0 = 0; z0 = 0;
X0 = [x0; y0; z0];
Q0 = eye(3); % orthonormal basis for tangent space
% The extended state vector contains [x; y; z; Q(:)]
Y0 = [X0; Q0(:)];

%% Initialization for accumulation of growth rates

sumLog = zeros(3,1);
count = 0;
t = 0;
Y = Y0;

%% Integration loop
for i = 1:N
% Integrate one step dt by RK4
Y = rk4_step(@(t, Y) extendedSystem(t, Y, m0, m1, alp, be), t, Y, dt);
t = t + dt;
% Re-orthonormalize every 'steps_per_orth' steps
if mod(i, steps_per_orth) == 0
Q = reshape(Y(4:12), 3, 3);
[Q, R] = qr(Q); % QR decomposition on tangent matrix
% Accumulate logarithms of the diagonal values of R (growth factors)
sumLog = sumLog + log(abs(diag(R)));
count = count + 1;
% Update extended state vector with re-orthonormalized Q
Y(4:12) = Q(:);
end
end

%% Calculate Lyapunov exponents (average over total duration)

LE = sumLog / (count * T_R);
fprintf('Calculated Lyapunov exponents:\n');
fprintf('LE1 = %f\n', LE(1));

63
fprintf('LE2 = %f\n', LE(2));
fprintf('LE3 = %f\n', LE(3));

%% Local functions

function dYdt = extendedSystem(~, Y, m0, m1, alp, be)

% Extract state variables
x = Y(1); y = Y(2); z = Y(3);
% Calculate h(x)
h = m1*x + 0.5*(m0 - m1)*(abs(x+1) - abs(x-1));
% Chua system equations
dx = alp*(y - x - h);
dy = x - y + z;
dz = -be*y;
% Derivative of h(x), piecewise definition
if x < -1
dh = m1;
elseif x <= 1
dh = m0;
else
dh = m1;
end
% Jacobian matrix J of partial derivatives
J = zeros(3,3);
% Derivative of (y - x - h(x)) w.r.t x is -(1 + h'(x))
J(1,1) = -alp*(1 + dh);
J(1,2) = alp;
J(1,3) = 0;
J(2,1) = 1;
J(2,2) = -1;
J(2,3) = 1;
J(3,1) = 0;
J(3,2) = -be;
J(3,3) = 0;
% Extract tangent matrix Q (from extended state)
Q = reshape(Y(4:12), 3, 3);
% Evolution of Q by variational equation Q' = J*Q
dQ = J*Q;
% Assemble derivative of extended state
dYdt = [dx; dy; dz; dQ(:)];
end

function Y_next = rk4_step(f, t, Y, dt)

% Runge-Kutta 4th order scheme
k1 = f(t, Y);
k2 = f(t + dt/2, Y + dt/2 * k1);
k3 = f(t + dt/2, Y + dt/2 * k2);
k4 = f(t + dt, Y + dt * k3);
Y_next = Y + dt/6 * (k1 + 2*k2 + 2*k3 + k4);
end

64